See

discussions, stats, and author profiles for this publication at:

https://www.researchgate.net/publication/223055639

Decolorization of Textile Industry

Wastewater by Photo Catalytic
Degradation Process

Article in Dyes and Pigments May 2001

DOI: 10.1016/S0143-7208(01)00014-6

CITATIONS READS

321 463

4 authors, including:

Orfan Zahraa
University of Lorraine
81 PUBLICATIONS 1,438 CITATIONS

SEE PROFILE

All content following this page was uploaded by Orfan Zahraa on 31 July 2015.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
 Dyes and Pigments 49 (2001) 117125
www.elsevier.com/locate/dyepig

Decolourization of textile industry wastewater by the

photocatalytic degradation process
C. Hachem 1, F. Bocquillon, O. Zahraa, M. Bouchy *
Departement de Chimie Physique des Reactions - UMIR 7630 CNRS ENSIC, 1, rue Grandville, BP 451,
F-54001 - Nancy Cedex, France

Received 24 October 2000; received in revised form 4 December 2000; accepted 15 March 2001

Abstract
The photocatalytic degradation of various dyes (Orange II, Orange G, Congo Red, Indigo Carmine, Crystal Violet,
Malachite Green, Remazol Blue and Methyl Yellow) has been studied, using P25 Degussa as catalyst. All dye solutions
underwent a decolourization. The kinetics of reaction have been studied and were found to be zero or rst order with
respect to the dyes. This was compared with the adsorption properties. The eect of the addition of hydrogen peroxide
has been studied. An enhancement of the rate has been observed in all cases and the order with respect to the additive
was found to be almost zero. It is dicult to give a general picture of the kinetics using these very dierent dyes but the
process was found to be eective for the decolourization of textile wastewater. # 2001 Published by Elsevier Science
Ltd.
Keywords: Dyes; Decolourization; Photocatalysis; Kinetics; Hydrogen peroxide; Textile industry

1. Introduction currently used, which work by direct precipitation

Textile industry wastewater is heavily charged
with unconsumed dyes, surfactants and sometimes
traces of metals. These euents cause a lot of
damage to the environment. In most countries
researchers are looking for appropriate treatments
in order to remove pollutants, impurities and to
obtain the decolourization of dyehouse euents
[13]. Various chemical and physical processes are
* Corresponding author.
E-mail addresses: michel.bouchy@ensic.inpl- nancy.fr (M.
Bouchy), orfan.zahraa@ensic.inpl- nancy.fr (O. Zahraa).
1
Present address: Chemistry Department, Faculty of Sci-
ences, Damascus University.
0143-7208/01/$ - see front matter # 2001 Published by Elsevier Science Ltd.
PII: S0143-7208(01)00014-6
and separation of pollutants [4], or elimination by
adsorption on activated carbon or similar materi-
als. In this case, the problem is only displaced, and
further treatments are indeed necessary in order to
separate the puried euents or to regenerate the
adsorbents; therefore, a new and dierent risk of
pollution is faced [5]. Alternatively, a photochemical
approach has been adopted. Indeed, ultraviolet
irradiation combined or not with oxidative agents
such as ozone or hydrogen peroxide leads to a com-
plete destruction of the pollutants [69], but the
presence of intermediates arising from the photo-
degradation reaction could be more harmful than
the pollutant itself [10,11]. Such inconvenience can
be avoided by using the photocatalytic degrada-
tion. In this process, hydroxyl radicals (OH
) are
118 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125
generated when the photocatalyst is illuminated in
the presence of water and air [1214], these ultra
reactive species associated with oxygen are able to
achieve a complete mineralization of organic pol-
lutants into carbon dioxide, water and other non-
toxic products [15,16].
The aim of this study is to analyse the kinetics of
decolourization of various types of dyes and to
assess the inuence additives have on the process.
2. Experimental procedures
2.1. Materials
The dyes Orange II, Orange G, Congo Red,
Indigo Carmine, Crystal Violet, Malachite Green,
Remazol Blue and Methyl Yellow were commercial
products purchased from Sigma Chemical Co., and
used without further purication (see Fig. 1).
Hydrogen peroxide (30%, w/w) was obtained from
Fig. 1. Developed formulae of various dyes used in this study.
Fluka. Puried water was obtained from an Elix
Millipore equipment. The titanium dioxide used,
TITANDIOXID P25 Degussa, was about 70%
anatase. According to the manufacturers speci-
cations [17], the elementary particle in dry powder
was approximately spherical in shape with a size
of 20 nm. The specic surface area, as measured
by N2 adsorption at 77 K, was 44 m2 g1 [18].
The experimental set-up for the photocatalytic
decolourization has been described previously [19].
The photocatalyst has been deposited on 3.848
cm2 glass plates from a suspension of TiO2.
2.2. Procedure
A suspension of commercial Degussa P25 TiO2
of 4 g l1 in pure water was prepared, the pH was
adjusted to about 3 and it was sonicated. A given
volume suspension was carefully poured on the
glass plate and allowed to dry out at 100
C for a
few hours.
 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125
After drying, the plate was red at 475
C for 4 h
(Fig. 2). During the heating, OH groups from the
catalyst surface and the support can react and lose a
molecule of water, creating an oxygen bridge, thus
increasing the adherence of the catalyst to the
support. Before deposition, the glass surface was
treated using a dilute acid solution of HF and
washed in a basic solution of NaOH in order to
increase the number of OH groups. The suspension
concentration 4 g l1 was chosen so as to get thin
enough deposits. This deposition process was car-
ried out four times in succession so as to increase
the total thickness. As shown in Fig. 3, the rst
coat does not cover all the surface but additional
coats lead to a complete coverage.
After each run, the catalyst was washed and
irradiated for several hours in the presence of H2O2
so that it was regenerated.
Electron microscopy was carried out on a Jeol
T33OA scanning microscope with the help of the
LSGC Laboratory. Prolometry measurements
were carried out on a UBM optical prolometer
equipped with a MICROFOCUS probe. Fig. 4
shows the topology of a TiO2 deposit which shows
a uniform surface of roughness of about 5 mm. In
addition, the average thickness was estimated to
be 10 mm [20]. BET areas were determined on a
FISON Sorptomatic 1900 sorptometer.
Photocatalytic measurements were carried out
on a homemade reactor designed for testing 3.8
48 cm2 at glass plates carrying the deposited lm
of catalyst (Fig. 5). The total volume of aqueous
Fig. 2. Procedure for the preparation of the TiO2 deposits.
119
solution was 0.2 l. The solution was continuously
circulated with a peristaltic pump, owing freely
over the catalytic surface and stirred in the 0.1 l
tank and could be considered as being constantly
air-equilibrated. The circulating ow was 0.4 l
min1.
As the extent of adsorption or photocatalytic
degradation was very small during a single pass
(0.5 min), the whole system could be considered as
a perfectly stirred sealed reactor with respect to
the solution. Irradiation of the photocatalyst was
carried out by a MAZDA-TFN18 18 W UV
uorescent lamp emitting around 365 nm and
positioned parallel to the plate. The average ux
on the plate was estimated to be about 2.103 E
min1 m2 [20]. Adsorption measurements were
made using the same set-up in the absence of
irradiation.
Absorption measurements to monitor the dye
concentration in the aqueous phase were carried
out on a Lambda-2 Perkin-Elmer UV/vis spectro-
photometer.
3. Results and discussion
The kinetics of photocatalytic degradation of the
various dyes are pictured in Fig. 6ag, as the
decrease in the relative concentration C/C
of the
dye with the time of irradiation, where C is the run-
ning concentration and C
the concentration at the
beginning of the irradiation. Note that the solution
120 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125
Fig. 3. Electron micrographs of P25 deposits (top: rst coat;
bottom: 4 coats).
was allowed to equilibrate with the catalyst prior to
irradiation so that C
is lower that the actual
initial concentration of the sample Ci. A typical
experiment is given in Fig. 7.
3.1. Adsorption on the catalyst
As a rule, adsorption was quite fast and the
equilibrium concentration C
was reached within
about 45 min. The contact was carried out up to 90
min, but no signicant change in the concentration
was observed.
Although the initial concentrations of the vari-
ous dyes were comparable, the equilibrium value
for C
varied indicating dierent adsorption
properties. The fraction of dye adsorbed was
expressed as A=(CiC
)/Ci as given in Table 1.
Fig. 4. Prologram of a TiO2 deposit.
Acid dyes, Orange II, Orange G, Remazol Blue
and Indigo Carmine exhibited a low adsorption
and in most cases A was not measurable (within
the experimental error of about 0.02). Alterna-
tively, cationic dyes, Malachite Green and Crystal
Violet were easily adsorbed. This dierence can be
attributed to a higher solvation energy for acid
dyes. The sulphonate group can indeed allow the
formation of hydrogen bonds contrary to the
quaternary amines of the cationic dyes [21]. As a
consequence, the equilibrium between the adsorbed
and the solvated forms is more in favour of the
adsorption for the cationic than the acid dyes. The
pH is about 5.5, and, therefore, is close to the iso-
electric pH of 5.8 [22] where the solid surface has a
neutral global charge; there is then little segrega-
tion according to the charge of the ions. The direct
dye Congo Red, which possesses sulphonate
groups, exhibited a high adsorption (A=0.1),
which can be attributed to both the high molecular
weight of this molecule compared to the other
acid, and the large hydrophobic middle section.
At a very acid pH, the acid dyes are in neutral
form, which is less hydrophilic than the ionic form,
and adsorb strongly. For example, at pH 1.3,
Orange II undergoes a fast and strong adsorption
leading to a visible colourization of the catalyst.
Under these conditions, practically no photo-
catalytic degradation was observed, which was
attributed to a complete coverage of the surface.
This indeed prohibits the adsorption of H2O or
OH on the surface, and inhibits the initial process
of formation of the reactive hydroxyl radicals.
 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125
Fig. 5. Experimental set-up for reactant adsorption and photocatalytic degradation.
3.2. Kinetics of photocatalytic degradation
The rate of photocatalytic degradation was
determined as the rate of disappearance of the
dye. This degradation also is the decolourization
process, as the possible intermediates formed did
not absorb in the visible range. The disappearance
was, therefore, followed by absorption in the visible
range. The rate of degradation was expressed as
the intrinsic rate relatively to the area of deposit
(i.e. the area of the glass support) under given
conditions of catalyst load (12.5 g m2) and intrin-
sic light ux (2.0 103 E min1 m2); the rate was
then expressed in units of mol min1 m2 deposit. This
is of practical interest in designing a photoreactor
as doubling the irradiated deposited area doubles
the rate of degradation in the reactor. Note that
this deposit area is not to be mistaken with the
much larger catalyst surface area. The kinetics
were found to be of order 0 or 1 with respect to
the dye. Plotting either C/C
or ln (C/C
) versus
time checked the validity of the order. Discriminat-
ing between the two orders requires an extent of
reaction higher than 0.3, which is not always the
case. When the extent was small, the order has been
assumed to be the same as in experiments carried
out in the presence of H2O2 as an accelerating agent.
It can be veried indeed in the case of fast
degradations that the order was not altered in the
presence of this agent.
121
These orders can be considered to be associated
with the limiting cases of the common Langmuir
Hinshelwood model. In this simple model the rate
r is expressed as [14,23]:
 
KC
r k
1 KC
where k is a rate constant (function of the light ux)
and K the adsorption constant. At low adsorption,
r is equal to kKC (order 1) and at high adsorption, r
is equal to k (order 0).
However, no correlation between the adsorption
and the order can be found in the results given in
Table 1. For example, low adsorption dyes,
Orange II and Orange G, have an order of 1 and 0
respectively. This can arise from the existence of
specic reactive sites on the surface; these could be
associated with ecient traps favouring the forma-
tion of a reactive hydroxyl radical. If these sites are
relatively few with respect to all the sites, the total
adsorption does not reect the adsorption on these
specic sites. Therefore, there may be a stronger
eective adsorption of Orange G compared to
Orange II. According to the LH model, the rate
would reach a maximum value at high adsorption
(order 0). However, the rate for Orange G is lower
than the rate for Orange II. This discrepancy could
be attributed to a dierent reactivity, which is
expressed by the rate constant k. The fragile group
122 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125

Fig. 6. (a) Photocatalytic degradation of Orange II following dark adsorption (&C
=9.9 105 mol l1, irradiation ux 7 107 E
min1); eect of H2O2 (&103 mol l1, * 102 mol l1, * 2 102 mol l1, ~ 5 102 mol l1, ~ 101 mol l1). (b) Photocatalytic
degradation of Orange G (C
=9.9 105 mol l1), same conditions as in Fig. 6a. (c) Photocatalytic degradation of Congo Red (C
=3.5
105 mol l1), same conditions as in Fig. 6a. (d) Photocatalytic degradation of Indigo Carmine (C
=9.5 105 mol l1), same condi-
tions as in Fig. 6a. (e) Photocatalytic degradation of Crystal Violet (C
=1.8 105 mol l1), same conditions as in Fig. 6a. (f) Photo-
catalytic degradation of Malachite Green (C
=1.9 105 mol l1), same conditions as in Fig. 6a. (g) Photocatalytic degradation of
Remazol Blue (C
=9.6 105 mol l1), same conditions as in Fig. 6a.
 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125 123

Cationic dyes exhibit a relatively high adsorp-

Fig. 7. Variation of the dye concentration vs. time, showing the
dark adsorption (negative time) followed by the irradiation
at time t=0 (Crystal Violet).
in these dyes is the NH group, which results from
an equilibrium between two tautomeric forms [21]
where an H atom is exchanged between O and N as
shown in Fig. 8. This equilibrium is much inu-
enced by the electron-attracting sulphonate
groups, which would result in an enhancement in
the amine form in the case of Orange II. Accord-
ingly, although the adsorption is lower for this dye,
a high reactivity leads to a higher rate of photo-
catalytic degradation compared with Orange G.
This shows the diculty in comparing the photo-
catalytic degradability of the various dyes.
tion but the rate of degradation is only twice that
of the low adsorbing Orange II. Moreover,
according to an order of the unity with respect to
the dye, the rate of degradation of the cationic
dyes would be higher at the initial concentration
of the acid dyes (104 mol l1). However, both
adsorption and reactivity play a role in the global
kinetics. It is probable that the azo group is more
likely to be attacked via the abstraction of H on
the NH form by the reactive species, assumed to
be the hydroxyl radical, than the phenyl groups
via addition. The reactivity of Remazol Blue can
also be attributed to the presence of an NH group.
The dispersed dye Methyl Yellow has a very low
solubility in water. Solvents such as methanol and
dimethylformamide have been used in order to get
a homogeneous solution. A very high rate of
decolourization was then observed, which suggests
a process involving the additional solvent, but this
was not further studied, as the addition of the
organic solvent cannot be envisaged in an indus-
trial process.
3.3. Limiting processes
It is of interest to compare the rate of adsorp-
tion ra with the rate of degradation r
. This rate ra

Table 1
Properties and kinetics parameters of various dyes

Dye Orange II Orange G Congo Indigo Crystal Malachite Remazol Methyl

Red Carmine Violet Green Blue Yellow

Class Acid Acid Direct Acid Cationic Cationic Reactive Dispersed

Type Azo Azo Diazo Carbonyl Triphenyl Triphenyl Quinon Azo
Color index 15,510 16,230 22,120 73,015 42,555 42,000 61,200 11,020
lmax (nm) 485 480 498.5 610 590 617 595 448
Initial concentration 8.9 9.9 3.9 9.7 2.0 2.0 9.6
Ci (mol l1)105
A <0.02 < 0.02 0.1 0.02 0.1 0.06 <0.02
Initial rate of deg radiation 1.07 0.42 0.38 1.84 1.90 1.92 1.04a 13.44b
r
(mol min1 m2)105
Order/dye 1 0 0 0 1 1 0
r
H2 O2/r
c 2.8 3.7 2 7 1.4 1.5 4.6
Order/H2O2 >0 0 >0 0 >0 >0 >0
" (l mol1 cm1)105 4.48 0.23 0.49 0.22 0.92 0.86 0.13
a
pH=2.85.
b
Average degradation rate over 150 min.
c
(H2O2)=102 mol l1.
124 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125

Fig. 8. Equilibrium between the two tautomeric forms in the acid dye Orange II.

can only be estimated if the amount of adsorption

is large enough to allow a measurable decrease in
the dye concentration in the aqueous solution.
This is the case of half the dyes only. As the reac-
tion of degradation follows the adsorption, the
initial rate r
is necessarily lower than the initial
rate ra
. This is eectively the case in Congo Red
where the ratio r
/ra
is equal to about 0.4 but the
situation is the reverse in the case of the cationic
dyes where r
/ra
is equal to about 2.5 and 6.5 for
Crystal Violet and Malachite Green, respectively.
This infers that the adsorption process involved in
the degradation is much faster than the adsorption
monitored prior to irradiation. This could be due
the existence of dierent types of sites as suggested
above or to an enhancement of the adsorption rate
in the presence of irradiation. Another important
process involved in the mechanism is the transport
process. This latter is very fast: as an example,
when a rapid adsorption takes place for Orange H
in acid medium, the initial rate is noticeably higher
than the largest degradation rate observed for the
most reactive dyes in the presence of H2O2.
3.4. Eect of H2O2
Usually this additive is known to increase the
rate of photocatalytic degradation by allowing an
enhancement in the quantum yield of formation of
hydroxyl radical [24,25]. This is observed indeed
for all dyes in the present study as shown in Fig. 9.
In most cases, a limiting rate is reached when
increasing the additive concentration, so that there
is little change between the addition of 102 or
101 mol l1. Addition of a concentration of 103
mol l1 is generally not sucient to increase the
rate signicantly. Consequently, the eect of the
additive has been given in Table 1 as the factor of
increase in the initial rate when using a con-
centration of 102 mol l1.
Fig. 9. Eect of the addition of H2O2 on the initial rate r
of
degradation of the dyes (conditions as in Table 1); * Orange
II, * Orange G, ^ Congo Red, ^ Indigo Carmine ( 1/2), ~
Crystal Violet, ~ Malachite Green, & Remazol Blue.
The picture above is a general trend. However,
some dyes exhibit a slight dierence. For example,
Congo Red, Orange II and Malachite Green do
exhibit a slightly positive order with respect to H2O2
in the range 102101 mol l1. The case of Orange
II is of special interest as the rate goes through a
maximum when increasing the concentration of the
additive. This eect could be due to a competition
for adsorption between the dye and the additive.
This is in agreement with a low adsorption of this
dye as suggested above.
Addition of H2O2 does not seem to alter the basic
kinetics, as the order with respect to the dye is not
modied. The eect could only be due to an increase
in the OH
radical formation. This suggests a similar
limiting value of the rate of decolourization of the
various dyes at high additive concentration. This
ideal case is not strictly observed. However, it can
be seen from Table 1 that the lower the intrinsic
rate, the higher it is increased by the additive.
Indigo Carmine is a noticeable exception to this
rule.
 C. Hachem et al. / Dyes and Pigments 49 (2001) 117125
4. Conclusion
The photocatalytic process has proved very
eective in decolourizing aqueous solutions of the
various textile dyes. As both adsorption and reac-
tivity play a role in the kinetics, it is dicult to get
a satisfactory general picture of the mechanism.
Comparing degradation and adsorption suggests
that the degradation takes place on specic types
of adsorption sites on the catalyst surface. The
order with respect to the dye is either zero or one,
which may be in relation with a limiting adsorp-
tion on the active sites. The addition of hydrogen
peroxide enhances the rate of degradation, which
is of much interest for improving the quantum yield
of degradation. Finally, it is to be noted that the
irradiation used falls within the UV range avail-
able from the sun, which suggests the design of a
solar photoreactor for a decolourization process in
regions where much solar energy is available.
References
[1] Galindo C, Jacques P, Kalt A. Photodegradation of the
aminobenzene acid orange 52 by three advanced oxidation
processes: UV/H2O2, UV/TiO2 and VIS/TiO2 compara-
tive mechanistic and kinetic investigations. J Photochem
Photobiol A: Chem 2000;130:3547.
[2] Akmehmet Balcioglu I, Arslan I. Treatment of textile
industry wastewater by enhanced photocatalytic oxidation
reaction. J Adv Oxid Technol 1999;4:18995.
[3] Porada T, Gade R, Fabler D, Gunther K. Quatum yield
of TiO2-photocatalysed degradation of Acid Orange 7. J
Adv Oxid Technol 1999;4:2038.
[4] Wu J, Eiteman MA, Law SE. Evaluation of membrane
ltration and ozonation processes for treatment of reac-
tive dye wastewater. J Environ Eng 1998;124:2727.
[5] Homann MR, Martin ST, Choi W, Bahnemann D.
Environmental applications of semiconductor photo-
catalysis. Chem Rev 1995;95:6996.
[6] Majcen-Le Marechal A, Slokar YM, Taufer T. Decolora-
tion of chlorotriazine reactive azo dyes with H2O2/UV.
Dyes and Pigments 1997;33:2818.
[7] Galindo C, Kalt Y, Taufer A. UV/H2O2 oxidation of
azodyes in aqueous media: evidence of a structure
degradability relationship. Dyes and Pigments 1999;42:
199207.
[8] Kuo W. Decolorizing dye wastewater with Fentons
reagent. Wat Res 1992;26:8816.
[9] Zhang F, Zhao J, Shen T, Hidaka H, Pelizetti E, Serpone
View publication stats
125
N. TiO2-assisted photodegradation of dye pollutants II.
Adsorption and degradation kinetics of eosin in TiO2 dis-
persions under visible light irradiation. Appl Cat B:
Environ 1998;15:14756.
[10] Anderson MA, Yamazaki-Nishida S, Cervera-March S.
In: Ollis DF, Al-Ekabi H, editors. Photocatalytic puri-
cation and treatment of water and air. Amsterdam:
Elsevier; 1993. p 40520.
[11] Larson SA, Falconer J. In: Ollis DF, Al-Ekabi H, editors.
Photocatalytic purication and treatment of water and air.
Amsterdam: Elsevier; 1993. p. 4739.
[12] Matthews RW. Photooxidation of organic impurities in
water using thin lms of titanium dioxide. J Phys Chem
1987;91:332833.
[13] Fox MA. Photocatalysis: decontamination by sunlight.
Chemtech 1992:68096.
[14] Chen HY, Zahraa O, Bouchy M, Thomas F, Bottero JY.
Adsorption properties of TiO2 related to the photo-
catalytic degradation of organic contaminants of water. J
Photochem Photobiol A: Chem 1995;85:17986.
[15] Tunesi S, Anderson M. Inuence of chemisorption on the
photodecomposition of salicylic acid and related com-
pounds using suspended TiO2 ceramic membrane. J Phys
Chem 1991;95:3399405.
[16] Matthews RW, McEvoy SR. A comparison of 254 nm and
350 nm excitation of TiO2 in simple photocatalytic reac-
tors. J Photochem Photobiol A: Chem 1992;66:35566.
[17] Degussa Corp. Technical bulletin pigments, No. 56. 5th
ed. Degussa AG, Frankfurt; 1990.
[18] Chen HY, Zahraa O, Bouchy M. Inhibition by inorganic
ions of the adsorption and the photocatalytic degradation
of organic contaminants in TiO2 aqueous suspension. J
Photochem Photobiol A: Chem 1997;108:3744.
[19] Zahraa O, Dorion C, Ould-Mame SM, Bouchy M.
Titanium dioxide deposit lms for photocatalytic studies
of water pollutants. J Adv Oxid Technol 1999;4:406.
[20] Ould-Mame SM. Etude de catalyseurs deposes pour reac-
teur photocatalytique de traitement des eaux. PhD thesis.
Nancy, INPL. 1998.
[21] Galindo C, Jacques P, Kalt A. Total mineralization of an
azo dye [Acid Orange 7] by UV/H2O2 oxidation. J Adv
Oxid Technol 1999;4:4007.
[22] Zahraa O, Chen HY, Bouchy M. Photocatalytic degrada-
tion of 1,2-dichloroethane on suspended TiO2. J Adv Oxid
Technol 1999;4:16773.
[23] Mills A, Davies R. In: Ollis DF, Al-Ekabi H, editors.
Photocatalytic purication and treatment of water and air.
Amsterdam: Elsevier; 1993. p. 595600.
[24] Salvador P, Decker F. On the generation of H2O2 during
water photoelectrolysis at n-TiO2. J Phys Chem 1984;
88:611620.
[25] Jenny B, Pichat P. Determination of the actual photo-
catalytic rate of H2O2 decomposition over suspended
TiO2. Fitting to the LangmuirHinshelwood form. Lang-
muir 1991;7:94754.